Fixed lexing of XML comments.

The previous setup would consume too much. For example the following HTML:

    <a><!--foo--><b><!--bar--></b></a>

would result in the following T_COMMENT token:

    "foo--><b><!--bar"

The new setup requires the marking of a start position. I'm not a huge fan of
this but there doesn't appear to be a way around this.

ext/java/org/liboga/xml/Lexer.rl

@@ -90,11 +90,12 @@ public class Lexer extends RubyObject
 
         byte[] data = rb_str.getBytes();
 
-        int ts  = 0;
-        int te  = 0;
-        int p   = 0;
-        int pe  = data.length;
-        int eof = data.length;
+        int ts   = 0;
+        int te   = 0;
+        int p    = 0;
+        int mark = 0;
+        int pe   = data.length;
+        int eof  = data.length;
 
         %% write exec;